ACCOUNTING QUALITY MODELS: A COMPREHENSIVE LITERATURE REVIEW

Transcription

1 International Journal of Economics, Commerce and Management United Kingdom Vol. III, Issue 5, May ISSN ACCOUNTING QUALITY MODELS: A COMPREHENSIVE LITERATURE REVIEW Cetin Yurt International BurchUniversity, Bosnia and Herzegovina Ugur Ergun Faculty of Economics and Social Sciences, International Burch University, Bosnia and Herzegovina ugur.ergun@ibu.edu.ba Abstract Radical changes of the global economy in the last years and the resulting developments in the financial economic theory caused the primary goal (i.e. profit maximization) to give its place to value maximization. Of course firms can have other goals but they are put in order as secondary, tertiary goals. In accordance with this goal, it is possible that managers manipulate the accounting numbers due to circumstances and incentives. Financial reports which have crucial information to their users like investors, employees, creditors, suppliers, customers, government and state deciding the issues about the firm s financial position and performance might be manipulated to achieve this primary goal. This paper analysis the models and techniques employed to detect manipulation in financial information presented in the financial statements. In overall, the related literature indicates that implementation of International Financial Reporting standards have superior impact and better reflect on accounting quality issue. Keywords: Financial statement manipulation, financial statement quality, financial reporting, quality measurement models, accounting quality Licensed under Creative Common Page 33

2 Cetin & Ugur INTRODUCTION Because of financial information manipulation, the real financial positions and operating results of firms cannot be reflected onto the financial information users. This on one hand causes the investors who invest in the securities of these firms to incur losses and lose confidence in the system, and on the other hand causes resources (funds) allocated to wrong and inefficient fields because of investment decisions in the firms and buying and selling decisions of investors about securities. As a result, additional costs are brought to the economy (Küçükkocaoğlu & Küçüksözen, 2005). When manipulation is in question, generally and normally most of the people think that firms decrease their incomes and increase their expenditures in order to pay less tax. However, the exact opposite of this is performed in order to achieve firm value (stock price) maximization which we briefly mentioned above. While firms apply earnings management, they choose to increase their incomes (for example making consignment sales look like normal sales) and decrease their expenditures (for example making sales and marketing expenditures look like research and development expenditures) by using accruals generally in the legislative framework. That is why, the results of a lot of studies reveal that firms increase their incomes by certain motives and incentives, and therefore implement earnings management. The frequency of the academic studies in recent years regarding the consistency of accounting figures and financial information manipulation has shown the necessity of these studies especially after Enron, WorldCom and Parmalat corporate scandals (Penman, 2003). At the heart of these studies which find a far-reaching place especially in the Anglo-Saxon literature the extent to which financial statements reflect the truth is observed (Küçükkocaoğlu & Küçüksözen, 2005). RELATED LITERATURE In the recent years, the importance of the academic studies that aim to determine accounting quality has become more understandable especially after accounting and auditing scandals. At the heart of these studies which find a far-reaching place especially in the Anglo-Saxon literature the extent to which financial statements reflect the truth is observed (Küçükkocaoğlu & Küçüksözen, 2005). In their decision making processes, investors who plan to make short or long term investments, credit agencies, suppliers and even employees wish to have information about the firm s performance. This demanded information being high quality is very important with regards to decisions because the accuracy of decisions hinge on high quality financial statements that can clearly and accurately express the financial position. The prominent determinants of quality here are relevance, accuracy, timeliness, accessibility-clarity and Licensed under Creative Common Page 34

3 International Journal of Economics, Commerce and Management, United Kingdom comparability in other words decision usefulness which is used very frequently in the literature. It seems possible to clearly describe the concept of quality in accounting only with the above-mentioned features. In addition, it is easy to observe in the literature that there is no unity of concepts and descriptions about this issue. Differences among countries like different financial systems and market structures, different development levels, different legal systems, different economic policies and accounting cultures which are the products of these, and most importantly different accounting and reporting standards can be counted as the obstacles of creating a single description of accounting quality. In addition to these, the decision makers who demand the financial data expectations from accounting quality can be different in line with their own evaluations and needs. The question decision usefulness for whom? widens the environment of the accounting quality concept. The existence of these differences and various approaches, the integration and globalization of capital markets, international trade reaching to a large scale, and incidents with similar features cannot remove the need to produce complete, needs-fulfilling and comparable information (İlker, 2010). The managers who would like to make the financial position of the firm look different from its original position (generally a better look is intended) with various motives and incentives, can have the tendency to manipulate the financial information both by using their right of choice that is presented to them or with different techniques and applications that can definitely be named as fraud. The need of investors, analysts, in short all the parties of the financial market is high quality accounting reports that have never exposed to any manipulations of the kind stated above and that gives clear information about the firm. The goal of the researches made and models created regarding accounting quality is to determine the extent to which financial statements reflect the truth. Despite the fact that the goal of the analyses and tests that are carried out to measure accounting quality are the same, different accounting items and environmental factors are used and that is why in the literature it is possible to encounter completely different models as well as a lot of models that are developed successively. In the literature, it is possible to encounter a lot of studies that are performed under various headings like accounting quality, earnings management, income smoothing, value relevance and fair value in general, and specifically creative accounting, numbers game, extreme accruals and accounting magic. Most accounting choice studies attempt to explain the choice of a single accounting method (e.g., the choice of depreciation) instead of the choice of combinations of accounting methods. Focusing on a single accounting method reduces the power of the tests since managers are concerned with how the combination of methods affects earnings instead of the effect on just one particular accounting method (Watts & Zimmerman, 1990). Some studies Licensed under Creative Common Page 35

4 Cetin & Ugur seek to explain accounting accruals (the difference between operating cash flows and earnings). Accounting accruals aggregate into a single measure and that is the net effect of all accounting choices (DeAngelo, 1986; Healy, 1985). Generally, studies on accruals use combinations of three sets of variables: representing the manager s incentives to choose the accounting method under bonus plans, debt contracts and the political process. Bonus plan and debt contract variables are used because they are observable. The three particular hypotheses most frequently tested are the bonus plan hypothesis, the debt/equity hypothesis, and the political cost hypothesis (Watts & Zimmerman, 1990). This section will continue with the detailed and comparative (with their advantages and disadvantages) examination of the models that are developed to measure accounting quality. As mentioned above, detailed information regarding the assumptions, the accounting items they contain and their ratios and their contributions to the accounting science of the models that come under three classes and that are due to different approaches will be given. ACCOUNTING QUALITY MODELS Models Based on Accruals The financial statements, which are organized to present the information that the demanders need on time, appropriately and accurately, are prepared on the accrual basis. Accrual in the accounting language is the recording of a financial event on time to the relevant account with regard to the periodicity principle regardless of cash inflow or outflow. According to accrual basis, the impact of transactions and other events are accrued to the relevant account not when cash or cash equivalents are collected or paid but when these transactions and events take place, and they are reported in the financial reports of that period (Örten, Kaval, & Karapınar, 2011). In other words, the fact that the results of the business financial transactions and events are reflected on financial statements without waiting for them to be converted to cash takes place by virtue of accruals. In order to be able to describe accruals in a more comprehensive way, Richardson, Sloan, Soliman, and Tuna (2005) state that in case there is no accrual based accounting, the only asset or liability item that will appear on the balance sheet will be the cash account. The reason for this is the fact that all other assets and liabilities are the result of accrual based accounting. In other words, assets other than cash are also included in the decision making process by using accruals. Thus accruals give the accounting incomes the ability to evaluate and measure performance (Durak, 2010). The fact that accruals are an important indicator of a firm s performance causes accruals to be a means to make this performance look different from its actual position by managers having this intention. Licensed under Creative Common Page 36

5 International Journal of Economics, Commerce and Management, United Kingdom Cash flows have impacts that are to reverse in the short run. For example managers can prefer to pay the debt due in the following period in order to write up the cash flows that are reported for one period; however the additive effect of cash flows this period can turn into a deductive effect in the following period. Accruals clean the accounting income from this kind of effects and provide it to be a more effective performance criterion (Ball & Shivakumar, 2006). Most of the models that constitute the accounting quality and earnings management theory which is the most important indicator of this quality center on accruals because in essence accrual (as a system) can be more easily managed as compared to profit and cash flows. Departing from this main thought, Jones (1991)brought forward the assumption earnings management will be put into practice not in the cash part of the profit but in the accrual part of it. This assumption has been tested by many researchers, its scope has been broadened and it has pioneered the formation of new assumptions and models. Since GAAP allows certain discretion to report accounting accruals, there is a possibility that accruals contain management s expectations about future cash flows or management s intention to manipulate information (Gomez, Okumura, & Kunimura, 2000). According to Dechow and Skinner (2000), as a natural result of accrual basis, managers have to carry out an evaluation and make a decision about the accrual time and amount of earnings and expenditures. The fact that this state is combined with various conditions and purposes inevitably causes profit management. In other words, the main reasons of earnings management applications are accrual accounting and the flexibility it provides to managers. With this flexibility provided, it is accepted that arrangements can be made on the earnings amount by using managers experiences and information advantages in order to estimate the future cash flows and remove the mistakes and deficits that the period s cash flows contain. In the methods regarding the examination of accruals, the profits of firms are divided into two components as follows: the profit that is composed of earnings that are collected as cash and paid expenses, and the profit component that is composed of accruals that have not yet been converted to cash. Since cash flow is independent from the accounting policies that managers pursue, the managers who want to make the profit look high will try to achieve their goal by applying methods that will make the amount of the accruals increase. According to Leuz, Nanda, and Wysocki (2003), when cash flows and accruals are compared in terms of information content, the result reveals that cash flows in firms contain more information. The reason for this is the opinion that as a result of accounting application alternatives in firms are more, the income smoothing incidents will be encountered more and that is why the earnings will become less related to value. That is why in researches; generally accruals are the point of departure in the determination and measurement of earnings management. Licensed under Creative Common Page 37

6 Cetin & Ugur Rajgopal and Venkatachalam (2008)argue that there is decrease of the rate of accruals turning into cash in the last 40 years in parallel with the decrease observed in the effectiveness of accounting information in terms of valuation. Based on this study and similar studies, it is possible to argue that accrual quality models can be utilized in order to research weather the reliability and therefore the effectiveness of accrual based accounting in reflecting firm performance decrease in time or not. The managers who want to have an advantageous position as compared to the parties who are interested in the firm s financial reports by making the firm look as though it had higher earnings will try to achieve their goals by applying methods about the amount and time of the accruals within the frame of the accounting policies they determine due to the fact that they cannot interfere in cash flows. Since the difference between accrual basis and cash basis is about time, there will be no difference between a firm s accruals and cash in the total operating period of the firm (Jones, 1991). However since earnings and expenditures are handled according to the periodicity principle in the short run, a difference will arise between the accrual basis and cash basis, and this state can be used on the way to earnings management. Subramanyam (1996) ask why do managers choose to manipulate accounting accruals? This is an important question in at least two respects. First, financial statement users are interested in how discretionary accruals should be interpreted in and what settings these numbers increase and decrease the informativeness of reported earnings. Second, standard setters tend to act to reduce managers' ability to exercise discretion in the reporting process, apparently based on the assumption that managers exercise their accounting discretion opportunistically. In the studies to determine and measure earnings management, the reasons to center upon accruals are as follows (Beneish, 2001): Accruals are the main product of GAAP and if the earning is to be managed, it will be accrued on the accrual side of the profit, not on the cash side, Centering upon the accruals decreases the problem in measuring the different accounting policy preferences that have an effect on profit, In case profit management is an unobservable part of the accruals, finding out the effect of profit management on the disclosed profit is improbable from the point of view of investors. While classifying accruals as accruals in current and fixed assets, and accruals regarding the main operation and finance, Richardson et al. (2005) made benefit of the comprehensive balance sheet approach. Teoh, Welch, and Wong (1998), classified accruals according to their Licensed under Creative Common Page 38

7 International Journal of Economics, Commerce and Management, United Kingdom terms as short and long term accruals. At the same time they classified accruals according to their state of being under the control of management as discretionary and nondiscretionary accruals. It should be noted that the short term and discretionary accruals are the accrual class that constitute the subject of the most number of studies that aim to determine short term accruals, long term accruals and earnings management. When the managers who can determine the accounting policy with different purposes and flexibilities that are presented to them want to manipulate the accounting numbers, they focus on the financial events that can be recorded in a discretionary way or can be left off the books. Discretionary accruals are the accruals that emerge depending on the discretional power or unexpectedly, and they are also called abnormal accruals or unexpected accruals. Examples of discretionary accruals can be allowance for doubtful receivables, worthless receivables, provisions no longer required, reorganization expenditures, effects of the changes in accounting estimations, profit/loss from sales of assets, accrued expenditures and deferred revenues (Bartov, Givoly, & Hayn, 2002).Nondiscretionary accruals are also named as normal accruals, and they are the accruals that are related to the firm s routine operations. As mentioned above, discretionary accruals are generally used as a measure and an indicator of managers accounting policy choices, and the degree of the discretionary accruals is estimated. In order to be able to reveal or measure discretionary accruals, the general starting point is the total accruals. It is observed that the models that are constituted to determine the accounting quality try to divide the total accruals into their discretionary and nondiscretionary components (Teoh, Welch, et al., 1998). In some studies, as a result of this distinction, the discretionary accrual amounts are indexed to the total assets or the sales revenue amount, and the tendency that comes into view in this index by years is accepted to be the indicator of financial information manipulation devoted to various purposes (Tekin & Kabadayı, 2011). According to Dechow, Sloan, and Sweeney (1995), most of the models require the estimation of at least one parameter and when no systematic earnings management is foreseen, this is typically provided by the usage of an estimation period. An equation that does not contain the accruals that are devoted to the analysis of financial statements could be in the form in which there was the comparison of the profit amount that would come into existence as a result of the accounting policies that the business administrators would choose in case they did not pursue the goal of running earnings management and the accounting policies that they would choose in case they gained favor as a result of earnings management. However, the facts that whether the accounting policies that business administrators apply in the relevant periods are applied with the intention of earnings management or not, which accounting policies they will apply in case no purpose comes into Licensed under Creative Common Page 39

8 Cetin & Ugur existence devoted to earnings management and the effects of these policies to the business profit cannot be completely observed by the researchers who are outsiders for the business, and that is why this method is not utilized in studies (Yükseltürk, 2006). In the literature total accruals (TA) are calculated in two ways: balance sheet-based approach and cash flow statement-based approach. In the studies that use the balance sheetbased approach (Healy, 1985; Jones, 1991), the total accruals are formulated as follows: TA τ = ΔCA τ ΔCash τ ΔCL τ ΔDCL τ DEP τ Where;ΔCA τ = Change in current assets in year t; ΔCash τ =Change in cash and cash equivalents in year t; ΔCL τ = Change in current liabilities in yeart; ΔDCL τ =Change in debt included in current liabilities in year t; DEP τ = Depreciation and amortization expense in year t. The second method that is used in the calculation of total accruals is the cash flow statementbased approach. According to this approach, total accruals are generally calculated as follows (Dechow et al., 1995): TA = NI CFO Where;NI = Net Income CFO = Cash from operating activities Hribar and Collins (2002)examined whether the balance sheet-based approach or direct calculation from the cash flow statement-based approach is more successful. The results of the study reveal that better results can be achieved by calculating the total accruals in the abovementioned way by making benefit of the cash flow statement under the same conditions, and the balance sheet-based approach gives inaccurate results in the estimation of the total accruals. In fact most of the studies that research the accounting quality made benefit of the cash flow statement figures in their models which can mean an indirect support for Hribar and Collins (2002). Zhang (2007) states that when accruals are handled as investments devoted to future periods, the accruals that are estimated by making benefit of the information that is obtained from the balance sheet are better growth criteria because these accruals both contain the organic growth in the working capital that the cash flow statement reflect and inorganic growth like mergers and acquisitions. To see how much of the accrued revenues and expenses that reveal the firm profit are concluded with cash and cash equivalent instruments of payment constitute the main goal of studies. The cash flow information derived from cash flow statements is the absolute must for such a control. Licensed under Creative Common Page 40

9 International Journal of Economics, Commerce and Management, United Kingdom HEALY MODEL (1985) Healy (1985) Model is the first model developed in the literature and estimates that the systematic earnings management will exist in every period. It is very simple and it is criticized as being quite insufficient in estimating discretionary accruals by researchers like Young (1999). In his study Healy (1985) tested the hypothesis that the managers who are given bonus schemes based on the firm s performance would want to increase the bonus schemes they would receive and apply earnings management. Healy (1985) stated that the firm s earnings are comprised of cash flows derived from operations, nondiscretionary accruals and discretionary accruals. The accounting transactions of nondiscretionary accruals and cash flows generally have the obligation to be carried out in the way that is determined by regulators like IASB, FASB or local standard setters that set the relevant rules. Discretionary accruals, on the other hand, can be recorded according to the accounting policies and methods that are determined by managers. The model assumes that managers within the frame of the provided opportunities can influence the earnings amount by periods by playing with discretionary accruals. This assumption departs from the point that this year s discretionary accruals are a component of last year s total accruals. Within the model, the discretionary accruals are expected to be zero. It is assumed that every firm having discretionary accruals other than zero applies profit management, every firm having discretionary accruals below zero operate in the direction of increasing profits and every firm having discretionary accruals above zero operate in the direction of decreasing profits (Aren, 2003). In addition, Healy (1985) states that because of reasons like limitations derived from the legislation and limiting factors derived from independent external audit, in case earning management towards increasing profits is applied in a certain period, an opposite policy should be pursued in the following period (Yaşar, 2011). In that case it is assumed that the total of earnings management applications will be zero during the time the managers are in charge. Healy (1985) explains three possible situations as follows by the assumption that the manager will make the discretionary accrual choice decision for two periods (current period and the following period) as long as s/he is in charge: In the first case the manager will want to choose the discretionary accruals that are in the direction of decreasing profit. There are two possibilities for this case. The first possibility is the case in which the profit before discretionary accruals is above the profit target that is determined for the bonus scheme (the lower bound). In that case the manager will not be able to exceed the bonus scheme lower bound through profit management applications and therefore will not be awarded by the bonus scheme; s/he will choose the way to minimize the discretionary accruals. The second possibility is the case in which the profit before discretionary Licensed under Creative Common Page 41

10 Cetin & Ugur accruals in the first period (period t)is within the discretionary accrual bound and within the bonus scheme lower bound. In that case the manager will either decrease or increase the discretionary accruals. If the manager chooses the way to increase the discretionary accruals, s/he will get bonus scheme in this period but s/he will give up the bonus scheme that is expected for the following period. If the manager chooses the way to decrease the discretionary accruals, s/he will maximize the bonus scheme that is expected for the following period but s/he will not be able to get bonus scheme in this period. Therefore the manager is in a position to make a decision between the present value of getting bonus scheme in this period and giving up the bonus scheme that is expected in the following period. In the second case, the manager will want to choose the discretionary accruals in the direction of increasing the profit. In that case it is assumed that the manager will understand that the lower bound determined to get the bonus scheme can be exceeded but the upper bound cannot be exceeded, and s/he will choose to apply discretionary accruals in the direction of increasing profits. In the third case the manager will want to choose the discretionary accruals in the direction of decreasing profits. In that case it is assumed that the manager will understand that the upper bound determined regarding getting bonus scheme cannot be exceeded and therefore s/he will use the discretionary accruals devoted to decrease the profit in this period with the purpose of increasing the following period s bonus scheme. Healy (1985) calculated the estimation of discretionary accruals that are used in earnings management by formulating his model as follows: NDA τ = TA t t T Where:NDA = Estimated nondiscretionary accruals; TA = Total accruals scaled by lagged total assets; t = 1, 2, T is a year subscript for years included in the estimation period; = a year subscript indicating a year in the event period. Holthausen, Larcker, and Sloan (1995) reexamined the extent to which earnings are manipulated to maximize the value of payments under short-term bonus plans. They found evidence like Healy, consistent with the hypothesis that managers manipulate earnings downwards when their bonuses are at their maximum. Unlike Healy, they found no evidence that managers manipulate earnings downwards when earnings are below the minimum necessary to receive any bonus. They demonstrate that Healy's results at the lower bound are likely to be induced by his methodology. Licensed under Creative Common Page 42

11 International Journal of Economics, Commerce and Management, United Kingdom DEANGELO MODEL (1986) DeAngelo (1986) tested the hypothesis that managers applied earnings management with the intention of making the stocks look less valuable while the publicly traded companies are brought to the non-public private company status with management buyout using the data of 64 companies that were traded between 1973 and 1982 in NYSE (New York Stock Exchange) and AMEX (American Stock Exchange). DeAngelo (1986) tests for earnings management by computing first differences in total accruals, and by assuming that the first differences have an expected value of zero under null hypothesis of no earnings management. This model uses last period s total accruals (scaled by lagged total assets) as the measure of nondiscretionary accruals. Thus, the DeAngelo Model for nondiscretionary accruals is: NDA τ = TA τ 1 The DeAngelo Model is considered a special version of the Healy Model (1985) in consequence of the facts that it does not require any estimation periods and the estimation period of the nondiscretionary accruals is limited by the previous year s observations (Dechow et al., 1995). Just like Healy, DeAngelo also accepts the fact that mathematically, discretionary accruals cannot be calculated alone(aren, 2003). A common feature of the Healy and DeAngelo Models is that they both use total accruals from the estimation period to proxy for expected nondiscretionary accruals. In case the nondiscretionary accruals are constant in the course of time and the discretionary accruals are zero in the estimation period, both Healy (1985) and DeAngelo (1986) Models will be able to measure the nondiscretionary accruals accurately. However, in case the nondiscretionary accruals change from period to period then both models will measure the nondiscretionary accruals inaccurately. In that case, the question which model is appropriate? will depend on the feature of the time series process that generates the nondiscretionary accruals. In case the nondiscretionary accruals follow a white noise process around a constant average then the Healy Model (1985) will be appropriate, and in case they follow a random walk process then the DeAngelo Model (1986) will be appropriate. Although the evidences regarding the facts that the total accruals are at a constant level and they are close to the white noise process reveal that the Healy Model (1985) will be more appropriate in measuring the discretionary accruals, opposite views are also asserted (Dechow et al., 1995). Both Healy (1985)and DeAngelo (1986)Models assume that the nondiscretionary accruals are constant in the time period examined. However this is not a powerful assumption (Dechow et al., 1995) because due to the nature of accrual-based accounting system, changes Licensed under Creative Common Page 43

12 Cetin & Ugur may occur in the level of nondiscretionary accruals with regard to the economic conditions of the firm (Kaplan, 1985). JONES MODEL (1991) Jones (1991) has brought a model to the literature in which the model itself confirms the assumption that nondiscretionary accruals are not constant. Unlike the Healy (1985) and DeAngelo (1986) models that contain the assumption that the average change in nondiscretionary accruals is constant and the change in total accruals stems from discretionary accruals Jones (1991)added the change in sales and the gross amount of fixed assets to the model in order to control the effects of the changes that may occur in the nondiscretionary accruals as a result of the firm s economic position. The Jones Model for nondiscretionary accruals in the event year is: NDA τ = α 1 (1/A τ 1 ) + α 2 (ΔREV τ ) + α 3 (PPE τ ) Where; ΔREV τ = revenues in year τ less revenues in year τ 1 scaled by total assets at τ 1; PPE τ = gross property plant and equipment in year τ scaled by total assets at τ 1; A τ 1 = total assets at τ 1; and α 1, α 2, α 3 = firm-specific parameters. Estimates of the firm-specific parametersα 1, α 2 and α 3 are generated using the following model in the estimation period: TA τ = a 1 (1/A τ 1 ) + a 2 (ΔREV τ ) + a 3 (PPE τ ) + υ τ, Where;a 1, a 2 and a 3 denote the Ordinary Least Squares(OLS) estimates of α 1, α 2 and α 3 and TA is total accruals scaled by lagged total assets. The results in Jones (1991) indicate that the model is successful at explaining around one quarter of the variation in total accruals (Dechow et al., 1995). The descriptive statistical analysis in the model of Jones (1991)is built on the expectation model that was used in the model of DeAngelo (1986). The DeAngelo Model (1986)takes into account the total accruals of the previous period as normal accruals (nondiscretionary accruals) as an indicator of the current period total accruals. The non-normal accruals (discretionary accruals) on the other hand are defined as the difference between current period total accruals and normal accruals. In this context, the assumptions the average change in nondiscretionary accruals is constant and the change in total accruals stems from discretionary accruals exist. In his model Jones applied negative and positive change, t- statistics and Wilcoxon Signed Rank Test for every variable. He calculated a scale by dividing the total assets of the previous period (τ 1) to the variables of the current period (τ). The main Licensed under Creative Common Page 44

13 International Journal of Economics, Commerce and Management, United Kingdom assumption in the Jones (1991) accrual model is that if there is a difference between the accruals of the current period and the previous period, the reason for that is the change in the discretionary accruals because nondiscretionary accruals do not reveal continuous change from period to period (Duman, 2010). With this model he developed, tried to determine whether the companies that would like to make benefit of import support (the increases in tariffs and quota discount) tend towards earnings management applications or not with the purpose of making their earnings look low in the auditing periods of the United States International Trade Commission (USITC). The study reached findings that reveal the fact that the executives of the companies who would like to make benefit of these tariffs conduct earnings management through discretionary accruals in the direction of decreasing profits. Jones (1991) Model controls abnormal firm performance. High quality earnings result from high quality accruals, and high quality accruals are nondiscretionary (Davis-Friday, 2010).On the other hand Bernard and Skinner (1996) and Healy (1996) state that the accruals that are considered as abnormal according to the Jones Model (1991) not only stem from the executives interventions but also can stem from wrong presentation. In addition, it is also possible that these accruals are abnormal by their nature even if they do not contain intervention. Bernard and Skinner (1996) express that although they do not contain interventions, abnormal earnings and expenditures are accepted as discretionary accruals in the Jones Model because, they cannot be associated to the change in earnings in a linear way. In other words, Jones (1991) assumes that earnings do not contain interventions. This means that interventions do not stem from earnings even though accruals contain interventions and therefore prejudice is created that earnings management is not conducted. Moreover, Defond and Park (2001) claim that the Jones Model ignore the periodical effects. That is why researchers implement a model that handles 3-month periods instead of the classic Jones Model while determining abnormal accruals. The Jones Model was improved by DeFond and Jiambalvo (1994) and Dechow et al. (1995), and took the form of a model that is widely used in the literature. DeFond and Jiambalvo (1994) contributed the model by stating that instead of commonly using the regression coefficients for every firm in sectors, calculating them separately for every sector will give better results. Dechow et al. (1995)added the change in receivables to the model and brought in the model to the literature. The new model is named the Modified Jones Model and has been widely used ever since it was developed. Hribar and Collins (2002) stated that the estimations derived from cash flows are more reliable and criticized the Jones Model because it does not use cash flows, and that can mean that the model may make the mistake of classifying the accrual items Licensed under Creative Common Page 45

14 Cetin & Ugur as they contain intervention even if they do not. Moreover, the thesis that firms with high earnings own discretionary accruals towards increasing incomes is not validated in the studies of Dechow et al. (1995). This reveals that earnings management can change depending on earnings or Jones Model (1991) may be defective. Other studies devoted to improve this model(for ex. Kothari, Leone, & Wasley, 2005; Teoh, Wong, & Rao, 1998) were done as well however these two models have been gained a more widespread acceptance and have been intensely used in the literature. THE INDUSTRY MODEL (1991) Another model considered in the literature is the Industry Model that is proposed by Dechow and Sloan (1991). The Industry Model relaxes the assumption that nondiscretionary accruals are constant over time. Instead of attempting to model the determinants of nondiscretionary accruals directly, the Industry Model assumes that the variation in the determinants of nondiscretionary accruals is common across firms in the same industry. The Industry Model for nondiscretionary accruals is: NDA τ = γ 1 + γ 2 median 1 (TA) τ Where; median 1 (TA) τ = the median vaule of total accruals scaled by lagged assets for all nonsample firms in the same 2-digit standart industrial classification (SIC) code.the firm-specific parameters γ 1 and γ 2 are estimated using OLS on the observations in the estimation period. The power of the Industry Model to decrease the measuring error in discretionary accruals depends on two factors that can be criticized. The first one is the fact that this model meets only the change in the nondiscretionary accruals that are common for the firms in the same sector. If the changes in nondiscretionary accruals to a great extent reflect the changes special to the conditions of the firm then the Industry Model will not be able to exclude the discretionary accrual indicators from nondiscretionary accruals. The second factor is the fact that the Industry Model presents the discretionary accruals that are interrelated among the firms in the same sector. This situation can cause a problem of the existence of profit management. The greatness of this problem depends on how related is the motive of profit management among the firms in the same sector (Dechow et al., 1995). MODIFIED JONES MODEL (1995) The Modified Jones Model is designed to eliminate the conjectured tendency of the Jones Model to measure discretionary accruals with error when discretion is exercised over revenue recognition. In the modified model, nondiscretionary accruals are estimated during the event year (i.e., the year in which earnings management is hypothesized) as: Licensed under Creative Common Page 46

15 International Journal of Economics, Commerce and Management, United Kingdom NDA τ = α 1 (1/A τ 1 ) + α 2 (ΔREV τ ΔREC τ ) + α 3 (PPE τ ), Where;ΔREC τ = net receivables in year τ less net receivables in year τ 1 scaled by total assets at τ 1. It is important to note that the estimates of α 1, α 1 and α 1 are those obtained from the original Jones Model, not from the modified model. The only adjustment relative to the original Jones Model is that the change in revenues is adjusted for the change in receivables in the event year (i.e., in the year earnings management is hypothesized). In the Jones Model, the whole sales earnings is considered as normal accruals and therefore it is assumed that no earnings entries are performed before the accrual conditions (for details see IAS-18) regarding the making of the entry come into existence completely. However, one of the earnings management techniques is the applications in which sales earnings are entered before they accrue. In case sales earnings are entered before they accrue, there will be increase in trade receivables, and in accruals as a result of this increase. That is why Dechow et al. (1995) saw the relevant deficiency of the Jones Model (1991) in the calculation of discretionary accruals, and developed the Modified Jones Model which is widely accepted in the literature. In other words, the difference of this model and its contribution to the science of accounting is the fact that it considers the assumption that the changes in the amount of sales on account may stem from earnings management applications (Yaşar, 2011). DECHOW AND DICHEV MODEL (2002) The models discussed until this point are models that are developed with the approach to estimating abnormal accruals. Dechow and Dichev (2002) developed a model that the origination and reversal of working capital accruals in a stylized firm. The model embodies the intuition that the timing of the firm s economic achievements and sacrifices often differs from the related cash flows and that the benefit of accruals is to adjust for the these cash flow timing problems. However, the model also reveals that the benefit of using accruals comes at the cost of including accruals components that initiate and correct estimation errors. The main starting point of this model is the possibility that the accruals and cash flows regarding financial events can occur at different times. The evaluation of the model regarding the accrual quality indicates whether the books are cooked or not with the intention of revealing a better performance regarding the timing and amount of the cash flows of accruals. In other words, while Dechow and Dichev (2002)evaluated the quality of accruals, they considered whether the accruals are turned into cash or not in the following year. To drive practical measures of working capital accrual quality, Dechow and Dichev (2002) use the following firm-level time-series regression: WC t = b 0 + b 1 CFO t 1 + b 2 CFO t + b 3 CFO t+1 + ε t Licensed under Creative Common Page 47

16 Cetin & Ugur Where; WC t = is the change in working capital from year (t 1) to (t). ( WC), is computed as, Accounts Receivable + Inventory Accounts Payable Taxes Payable + Other Assets (net) CFO t 1 = the cash flows that created cash flows in the previous period but the effect of them on the earnings took place in the period (t), CFO t = the cash flows that both create cash flows and affect the earnings in the period (t), CFO t+1 = the cash flows that affect the earnings in the period (t) although they will create cash flows in the following period, ε t = represents accruals that are not turned into cash and their standard deviation is considered as the measure of the firm s accrual quality. This model uses total cash flow from operations (CFOs). Thus, the independent variables in expression are measured with error, implying that the regression coefficients are likely to be biased toward 0, and the R 2 will be reduced. Theoretically, the coefficients are expected to give the values0 < b 1 < 1 and -1 <b 2 < 0 and0 < b 3 < 1. Working capital accruals and cash flows that stem from main activities are used in order to measure the accrual quality. The reason to use the working capital accruals is the fact that it is known that they will turn into cash in one year. The error term that comes in view in the relationship between the change in the working capital and the cash flows that stem from main activities that belong to the previous period, the current period and the following period expresses the accrual estimation error. Error terms are not about cash flows. These terms contain estimation error and their standard deviations constitute a measure of accrual quality. In this model in which the factors that decrease the quality of accruals in businesses are handled it is stated that the size of the firm is directly proportional to the accrual quality. The results of the above-mentioned study reveal that the accrual quality is higher in large businesses. Dechow and Dichev (2002) claim that the businesses in which the accrual quality is lower own more accruals that do not turn into cash, and that means that their earnings are less perpetual. Dechow and Dichev (2002)attribute the fact that accruals are erroneously reported to the opportunist behaviors of executives and various other features special to the business like fluctuations in the activities of the business and the lengthiness of the operation period. For example if the activities of the business reveal extreme fluctuations, it is possible that the executives will make bigger mistakes in estimating accruals even if they are quite talented and well-intentioned. In the Dechow and Dichev Model (2002), it is thought that the error terms are independent from each other and independent from the cash flows accrued because in this model the intervention of management is not considered. In the literature, on the other hand, the Licensed under Creative Common Page 48

17 International Journal of Economics, Commerce and Management, United Kingdom opposite opinion is defended and it is stated that the errors that stem from earnings management are not dependent from each other and the accrued cash flows. In that case this model is not beneficial in measuring the quality of accruals in the financial statements that contain executives interventions. It is possible that the model will give wrong results with the classification error named Type I and Type II. McNichols MODEL (2002) McNichols (2002), criticized the fact that the Dechow and Dichev (2002) Models do not handle accruals as discretionary and nondiscretionary accruals, and stated that this distinction that is used in the Jones Model (1991) should be used in a model to be developed. McNichols (2002) also stated that in the Jones Model (1991), the working capital accruals are only influenced from the change in that year s sales and the changes of the previous or following year is ignored. In other words, McNichols claim that this model is not a sufficient model alone in evaluating the accrual quality because of the fact that it evaluates every period in itself, and it does not consider the following and previous periods. As a result, McNichols (2002) assumes that combining the two models will increase the explanatoriness of both models and will make the two models reduce the mistakes of each other. The estimation results in the Dechow and Dichev Model (2002) reveal that the adding of cash flows to Jones (1991) Model will decrease the model s skipping of the variables regarding economic events. In that context, McNichols (2002) also expressed that the measurement errors in the Dechow and Dichev Model (2002) may hinder the effective control of the basic variables that affect the accruals. That is why adding the earnings variable to the Dechow and Dichev Model (2002) consists a useful check point with regards to measuring the error in cash flow values. In that study, the fact that whether cash flows have explanatoriness over accruals or not after the changes in fixed assets and earnings considered is put forward with the following model: WC t = b 0 + b 1 CFO t 1 + b 2 CFO t + b 3 CFO t+1 + b 4 ΔSales t + b 5 PPE t + ε t Where;ΔSales t = the change in sales PPE t = Gross property plant and equipment in year τ scaled by total assets at t 1 In his study, McNichols (2002) states that both models are described in a wrong way, and sales and cash flows have high relationship. Then the result that the error terms are in high correlation with the variables in the other model comes in sight. The fact that these variables are related more to the error term than intervention reveals that some accruals that were estimated to contain interventions in the Jones Model (1991) do not contain interventions. Licensed under Creative Common Page 49

18 Cetin & Ugur LARCKER AND RICHARDSON MODEL (2004) Larcker and Richardson (2004),added the book-to-market ratio (BM) and cash flow from operations (CFO) to the Modified Jones Model to mitigate measurement errors associated with discretionary accruals. BM controls for expected growth in operation, while CFO controls for current operating performance. Previous research has shown that measures of unexpected accruals are more likely to be misspecified for firms with extreme levels of performance (Dechow et al., 1995). Accordingly, Larcker and Richardson (2004) include current operating cash flows, CFO, as an additional independent variable. Larcker and Richardson (2004) argue that their model outperforms the modified Jones model. Since discretionary accruals equal total accruals minus estimated non-discretionary accruals, the estimated non-discretionary accruals of the cross-sectional modified Jones model with book-to-market ratio and cash flow from operations are estimated as follows: TA t = α + β 1 ΔSales t ΔREC t + β 2 PPE t + β 3 BM t + β 4 CFO t + ε Where;BM t = is the market value / book value in year τ scaled by total assets at t 1 All variables are scaled by the average of total assets using assets from the start and end of the fiscal year. FRANCIS et al. MODEL (2005) Francis, LaFond, Olsson, and Schipper (2005) applied methods based on the modified Jones model approach and the Dechow and Dichev Model approach to separate either total accruals or current accruals into normal components (i.e., the portion associated with accounting fundamentals) and abnormal components (i.e., accruals that are not statistically associated with accounting fundamentals). Under the modified Jones model approach, accounting fundamentals are revenues adjusted for receivables and gross property plant and equipment (PPE). Under the Dechow-Dichev Model approach, lagged, current and lead cash flows from operations are the accounting fundamentals. Their seven earnings quality metrics capture various aspects of the abnormal component of accruals; the weaker the association between accruals and accounting fundamentals, the lower is earnings quality. Francis et al. (2005) modify and extend the Dechow and Dichev (2002) model in two ways. First, as suggested by McNichols (2002), they add growth in revenue in an attempt to reflect performance, and they add gross property plant and equipment, which expands the model to a broader measure of accruals that includes depreciation. However, Francis et al. (2005) do not investigate whether these adjustments help or hinder Type I or Type II misclassification errors. The second way they extend the Dechow and Dichev model is to decompose the standard deviation of the residual into firm-level measures of innate estimation Licensed under Creative Common Page 50